264 The Roman Villa at Boz. 
There was apparently nothing in the way of veining or “ marbling,” 
but the effect was produced solely by larger or smaller splashings 
ov spirtlings of different colours from a brush. Specimens of the 
following were found and have been set in plaster frames and 
preserved in the Museum :— 
A dark greyish green ground covered thickly with minute 
spirtles of dark red and white. This gives exactly the effect 
of a fine-grained granite or serpentine. This seemed the 
commonest of the “marble” colourings. 
A dark red finely spirtled with white, the effect being that of 
fine-grained red granite. This and the preceding seem to 
have been used as panels next to each other and divided only 
by a narrow white line. 
A lighter pinkish red ground with rather large splashes of white, 
and more sparingly of dark red. The effect is that of a large- 
grained pink granite. Very little of this was found. 
A white ground, splashed sparingly with rather large blood-red 
spots. Only a few pieces of this were found. 
A white ground, with very small spirtles of dark red, very 
sparingly applied. This, also, was scarce. 
Several fragments of apparently figures or pictures were found, 
nearly all having the subject painted in a dark purplish maroon, on 
a light green ground. This maroon colour was in worse condition 
and scaled off more easily than any other. 
Some large pieces of white plaster were found, but whether they 
were originally white, or whether the colours painted on the white 
ground had washed off, was not evident. 
A good many fragments of a solid plain quarter-circle moulding 
in plaster all coloured a purplish maroon occurred, and in one 
case a piece was found still in situ, showing that this moulding 
ran round the edge of the tessellated floors at the junction with 
the walls. Similar plaster “skirtings” have been found in other 
Roman villas.!| One or two other bits of moulded plaster, Boseielsl 
portions of door jambs, were found. 

1 See an example in the villa at Wheatley (Oxon), illustrated in Arche@o- 
logical Journal, II., 353. 

